Austin’s initial pitch: Create a engine building game that can be played by 2-4 players in 30-90 minutes, which is actually fun and also teaches some aspect of AI race dynamics

Questions

• Who is this for?

  • gamers?
  • ingroup?
  • decisionmakers
  • general public?

• Why are we doing this?

  • What is the experience we want players to have?
  • What is the message we want players to take away?

• What kinds of games are we basing this on?

  • Which board games do a good job of teaching a model? of inspiring action?

• Do we want to

  • Simply reskin a game?
  • Clone mechanics?
  • Design our own?

• Is a board game or an online game form factor better?

  • Pros of board game
    • More awesome
    • Effect of “let’s sit down and play this together”
  • Pros of online game
    • Cheap distribution, more virality potential
    • Can integrate LLMs. (though: hybrid board/online games?)
    • Easier to update

• How do we trade off simulation fidelity vs “fun”?

Game design parameters

Scoring

• Hidden and public values
• Could be independent scoring vs classic “winner takes all”